Digital Storage Oscilloscope 16
4.
Turn the VOLTS/DIV knob to ensure the entire waveform remains on the screen. If
the entire waveform is invisible, the oscilloscope may display wrong FFT results
by adding high-frequency components.
5.
Turn the SEC/DIV knob to provide the resolution you need in the FFT spectrum.
6.
If possible, set the oscilloscope to display multiple signal cycles.
If you turn the SEC/DIV knob to select a faster setting (fewer cycles), the FFT spectrum
will display a larger frequency range and reduce the possibility of FFT aliasing.
To set the FFT display, follow the steps below.
1.
Push the MATH MENU button;
2.
Set the Operation option to FFT;
3.
Select the Math FFT Source channel.
In many situations, the oscilloscope can also generate a useful FFT spectrum despite
the YT waveform not being triggered. This is especially true if the signal is periodic or
random (such as noise).
Note: You should trigger and position transient or burst waveforms as close as possible
to the screen center.
Nyquist Frequency
The highest frequency that any real-time digital oscilloscope can measure without
errors is half of the sample rate, which is called the Nyquist frequency. Frequency
information beyond the Nyquist frequency is undersampled which brings about the FFT
aliasing. The math function can convert the center 2048 points of the time-domain
waveform to an FFT spectrum. The resulting FFT spectrum contains 1024 points from
DC (0Hz) to the Nyquist frequency. Usually, the screen compresses the FFT spectrum
horizontally to 250 points, but you can use the FFT Zoom function to expand the FFT
spectrum so that you can clearly view the frequency components at each of the 1024
data points in the FFT spectrum.
Note: The oscilloscope
’s vertical response is a little bit larger than its bandwidth
(60MHz, 70MHz, 100MHz or 200MHz, depending on the model; or 20MHz when the
Bandwidth Limit option is set to Limited). Therefore, the FFT spectrum can display valid
frequency information above the oscilloscope bandwidth. However, the amplitude
information near or above the bandwidth will not be accurate.
